August 13, 1920] 



SCIENCE 



147 



segregate Pica Brisson, as the type and only 

 genus of a separate family, the name of such 

 family could not well be Picidee, since this is 

 already in use for another group, with Picits 

 Linnaeus as basis. Consequently the name of 

 the family containing Pica would become 

 Propicidw. 



Harry C. Oberholser 

 U. S. Biological Survey 



FURTHER RESULTS OF ANALYSIS OF 



LIGHT DEFLECTIONS OBSERVED 



DURING SOLAR ECLIPSE OF 



MAY 29, 1919 



1. Since the article in Science of June 11, 

 1919 (pages 581-585) was written, we have 

 received through the kindness of the Astron- 

 omer Royal the printed " Report "^ giving in 

 detail the reductions and results of the light 

 deflections observed by the two British expe- 

 ditions during the solar eclipse of May 29, 

 1919. On the basis of the information in 

 the " Report " we have made an independent 

 reduction of the photographic measures re- 

 sulting from Crommelin's plates. 



The non-radial effects, as resulting from 

 our calculations, are found to be on the 

 average about one third of those derived 

 from the British printed results and as 

 given in the seventh column of Table II. 

 of the previous article in Science (see page 

 583) ; in brief, our non-radial effects are on 

 the order of the error of observation, so that 

 they may he regarded as non-existent until 

 other observational evidence is obtained. 



2. Table I. contains the revised radial light 

 deflections resulting from all reductions; 

 they are subject to some slight changes when 

 some required additional information has 

 been received. Comparing the observed de- 

 flections with those computed on the basis 



i"A Determination of the Deflection of Light 

 by the Sun 's Gravitational Field from Observations 

 made at the Total Eclipse of May 29, 1919, ' ' by 

 Sir F. W. Dyson, F.E.S., astronomer royal; Pro- 

 fessor A. S. Eddington, F.R.S., and Mr. C. David- 

 Bon, Phil. Trans. B. S., London, Ser. A., Vol. 220, 

 pp. 291-333. [The longitude of Sobral, as given 

 on page 296, should read 2'' il"" 25' west, instead of 

 2" 47"' 25'.] 



of the Einstein theory of gravitation, it will 

 be seen that generally the observed de- 

 flection is greater than the theoretical value. 



TABLE I 



Sadial Light Deflections, May £9, 1919, at Soiral 



Star 11, the most distant star, accord- 

 ing to the British reductions showed a de- 

 flection agreeing better with the value cal- 

 culated on the basis of the Newtonian Me- 

 chanics, but it now shows a deflection agree- 

 ing better with the Einstein value. In brief, 

 the results of all reductions would lend addi- 

 tional support to the conclusion reached by 

 the British astronomers, namely, that, as 

 judged by their best photographic plates, the 

 light deflections observed during the solar 

 eclipse of May 29, 1919, accorded better with 

 the calculated values on the basis of the Ein- 

 stein theory than on the basis of the New- 

 tonian Mechanics. 



3. Comparing the observed deflections with 

 the theoretical ones, as given in Table I., it 

 would seem that the former decrease with 

 distance more rapidly than do the latter. 

 Whether this implies that the observed light 

 deflections were the combined effects of the 

 sun's gravitational action and a solar atmos- 

 pheric action of some kind can possibly not 

 be settled definitely until further observa- 

 tional evidence has been obtained.^ 



Louis A. Bauer 



Department of Terrestrial Magnetism, 

 Washington, D. C, 

 July 17, 1920 



2 Expressed in units of the sun 's radius. 



8 It may be suggestive that the light ray from 

 star 2, which according to Table I. differed largely 

 from the Einstein value, passed through the solar 

 atmospheirdc region directly above the remarkable 

 prominence on the southeast limb of the sun. 



